Productivity and time management

How to prioritise at work

BGL Tech’s Stephen Ramsey reveals how to order your to-do list and be more productive in three simple steps — even when everything needs doing ‘right now’:

This is the story of how the unprecedented events of COVID-19 led to a fundamental change to one of our key processes - prioritisation. We thought we had it pretty much nailed, and to be fair, in the normal day-to-day operation of software delivery, we probably weren’t doing too badly, but there’s nothing like a global pandemic to demonstrate that you could be doing things very, very differently.

Deciding what not to do

Like any normal software development house, we have more change in our combined backlog than we have capacity to get to. This is perfectly natural, and I’d argue that if your team doesn’t have more work than capacity, then your team is too big.

That means though, that you’re going to need to make choices; choices not just about what to do first, but about what to do at all. That is what prioritisation is; a decision about what to do, or perhaps more importantly, about what not to do.

So how do you make that decision? Well here’s how we were doing it.

Firstly, we viewed the process of prioritisation as one of comparison. In order to know what to do, and what not to do, you need to be able to compare one piece of work against another. In turn, to do that, you need to decide on some criteria that will apply to all of your potential changes, that you can then compare. Here’s what we came up with:

• Estimated size
  There’s a school of thought that says that estimates are counter-productive, and there’s a good argument for that with larger projects, but the majority of our change takes three months or less to deliver, so estimation, in ranges of sprints is still something we do. Everything that we put into the backlog for prioritisation has an estimated size.
• Estimated value
  Everyone knows you want to release value early right? That means you need to understand the value any particular change delivers. Every change going into the backlog for prioritisation had a value expressed as increase in revenue over the 12 month period from launch. Combined with the estimated size, this gives a return on investment (ROI) that you can compare between changes.
• Confidence in value
  Salesmen over-promise right? Well shock, horror, it’s the same in business modelling. Sometimes the projected value of a change is little more than a guess. As an organisation we still wanted to make those guesses though, so to inform prioritisation we put a confidence factor on the value. It’s still potentially worth doing a change that has a 30% chance of generating a 500% ROI before a change that has a 100% change of generating a 120% ROI.
• Contractual commitment
  Have we promised this to anyone? Even if it’s big, even if it doesn’t have a huge ROI, if we’ve committed to something in a contract, then up the prioritisation list it goes.

For each of these factors, we gave all changes a score. Once every three months when we were doing our next planning increment, we’d get together and look at the numbers.

Evaluate and schedule

We had distilled order from chaos. But even then though we weren’t done. Anyone can use a spreadsheet right? We were more subtle than that, and the job of the prioritisation group, made up of directors of all the business areas driving the change, and the delivery managers of the technology teams who would deliver it, was to apply the art to the science.

We looked at what the numbers told us as a first draft, and then we looked at the subtleties that were too fine to accommodate in a simple Excel model. OK, it’s a contractual commitment, but does the partner really care about that change, are they even going to be ready? OK, so change A has a higher ROI than change B, but change B might unlock value for a new sale coming along later.

And then even when you have an ordered list, you get into scheduling, which — and don’t let anyone tell you different — is still prioritisation. Where are the skills available for this change ahead of that one? We have a natural “one sprint” gap here, what can we pull forward?

We had a slick process, right from estimation through to execution, and we thought it was pretty robust. We congratulated ourselves on really understanding prioritisation, and in delivering value and all the other things that agile textbooks say you should be doing.

And then COVID-19 happens

Like almost everyone else, I was taken by surprise how quickly the coronavirus situation escalated. In truth, I think previous outbreaks like SARS, Swine Flu, Avian Flu, and Zeeka and perhaps to some extent even Foot and Mouth — none of which had disrupted society in the UK to a great extent, had lulled us all into a state of complacency. Even as cases were increasing in China, I don’t think we understood the full force of what was coming our way.

And then almost from nothing, to lockdown. Everything closed. Everyone who could, working from home. No toilet roll or pasta, even once you got through the queue to get into the supermarket. Strange times indeed.

We needed to get all our employees working from home. Most of us to be fair, just picked up our laptops and went, but for our contact centre staff, life wasn’t that straight forward. We had to provide the hardware for them to work on at home, and upgrade the software to allow them to do so.

Of course, with such a radical shift in the societal context, came a shift in customer behaviour, and in turn, in the required changes coming through our backlog. With a whole raft of changes starting to be presented by the business areas who wanted to ensure their customers were protected:

· Don’t charge fees for changes to policy for customers facing financial difficulty as a result of coronavirus

· Allow customers to add charity work to their acceptable use on their policy for free if it was for COVID-related activity

· Stop selling products that customers couldn’t utilise given they were in lockdown

The list went on and on.

How to prioritise when everything needs to be done NOW

And this is where our previously slick and robust prioritisation process failed us completely. For a start, we didn’t have the luxury of time. We’d been used to prioritisation on a three-month horizon, meaning that in January, we’d prioritised work we’d do between April and June. All this stuff needed doing NOW!

We could still provide an estimate, but what about the value? How do you place a value on your colleague’s health and wellbeing? Spoiler alert - you can’t. How do you place a value on treating your customers fairly and supporting a national effort? You still can’t.

Without value, there’s no ROI, and without that, there’s no basis for comparison, and as we all know, prioritisation IS comparison.

A new approach

Well the truth is, we went from making judgements about value, to making value judgements.

Get everyone working from home. Priority one. Good for the colleagues, fantastic for customers who can still access a great level of service.

Beyond that we did what felt right, but we broadly followed these principles:

• Colleague and customer welfare before commercial ROI
• Get as many people involved in the decision as is practical
• Be prepared to change your mind

No Excel spreadsheet, no score, just a really good conversation with people representing all the different aspects of delivery, and perhaps the most fundamental aligning principle was a shared goal. Sure, changes were in competition with each other, but actually they were all competing to perform against the same desired outcome; respond to COVID. No longer was sales penetration battling against a new product launch, or brand roll-out of some function going up against renewal rate. It seemed every change was pulling in the same direction, and how well it performed, however qualitatively that was judged, became the basis of comparison.

Honestly, I enjoyed it. I spent a lot of my working life in a ‘helpdesk’ context, so responding to things going wrong is something I’m used to. It is of course not something I relish.

So has COVID change how we prioritise? For a while it did, absolutely. When we were in the first shock of response all the rules seemed to go out of the window. But you can get toilet roll and pasta again now, and we’re coming into a ‘new normal’, so we’re starting to get back to a more regular routine. The methodical, numeric, almost scientific approach to prioritisation has returned. Now though there’s a new understanding — as if we ever really needed reminding — that conversation, exchanges of views and human value judgements are every bit as important, and do you know what, I really like that.